This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Monte Carlo simulation of presynaptic calcium dynamics and neurotransmitter release. This computational project is directed by J. Stiles and is being carried out by John Pattillo, a post-doc in his lab. Using realistic nerve terminal ultrastructure and data such as that described in II.A.1.c, MCell simulations of active zone calcium dynamics encompass action potential-activation of voltage-gated calcium channels, stochastic calcium ion entry and diffusion, calcium binding to sensor sites on arrays of synaptic vesicles, and prediction of vesicle fusion and resulting transmitter release. To our knowledge, this is the only study to date that has included the 3-D structure of an entire presynaptic active zone, and that has used multiple experimental constraints to enable quantitative predictions, e.g., the number of calcium-binding sites on synaptic vesicles, and the relationship between number of binding sites, number of sites that must be bound to initiate neurotransmitter release, and the importance of active zone spatial organization. In brief, a supralinear (~4th order)1 relationship (CRR) between extracellular Ca2+ ([Ca2+]o) and transmitter release indicates that multiple Ca2+ ions are required for fusion of a synaptic vesicle (SV), but how this empirical observation relates to the stoichiometry and architecture of voltage-gated Ca2+ channels (VGCCs), Ca2+ binding sites, and SVs is unclear. We created a spatially realistic model of a frog neuromuscular active zone (AZ), and used MCell to simulate action potential (AP)-induced Ca2+ influx through VGCCs, Ca2+ binding to SVs, and several models of Ca2+-dependent SV fusion. We varied spatial parameters to simultaneously reproduce 3 experimental observations: 1.) average release probability (pr) per trial per AZ at physiological [Ca2+]o; 2.) the distribution of release latencies (Ldis);and 3.) the 4th order CRR. Also, a 4-state VGCC model reproduced macroscopic Ca2+ current kinetics, and the on and off rates for Ca2+ binding were based on the synaptotagmin-1 C2A domain. Given all these constraints, we obtained a surprisingly unique set of model parameters and several counter-intuitive predictions. With a VGCC:SV stoichiometry of 1:1 (supported by the experimental and mathematical modeling data outlined above), each SV contains ~20 Ca2+ binding sites, and 6 sites must be bound simultaneously to induce fusion. Alternative models were either much too Ca2+-insensitive to reproduce pr or could not simultaneously reproduce Ldis and CRR. These results demonstrate the dramatic sensitivity of CRR, pr, and Ldis to presynaptic architecture, and suggest that vesicle fusion may require a variety of SNARE protein and membrane lipid binding sites for Ca2+. This work has been published in abstract form (Pattillo et al., 2004), and several full length manuscripts are in preparation. This project has required something on the scale of 105 simulations to date, primarily run on the PSC HP GS1280 machine(s), for which we are one of the preferred user groups. This machine is based on latest-generation Alpha EV7 processors, large shared memory, and outstanding memory bandwidth, and is optimally suited to our Monte Carlo algorithms and run-time optimizations within MCell. Specifically, MCell simulations require larges amounts of memory with random access patterns. In addition, this project admirably demonstrates the advantages to MCell's unique Monte Carlo algorithms for bimolecular interactions. The spatial dimensions of the active zone are tightly confined, and our simulations show that the average calcium concentration in the vicinity of vesicular binding sites corresponds to less than a single ion at any instant in time. Despite these conditions, MCell is able to accurately simulate these calcium dynamics with a time step on the sub-microsecond scale, rather than the sub-nanosecond scale (as would be required with less sophisticated algorithms for bimolecular interactions). Thus, this project has been possible only through a combination of optimized algorithms coupled with outstandingly designed and supported hardware. Computational Challenges These simulation have been performed using PSC's Marvel systems. Within this study, we are usually running one "project" at any given time. Each "project" includes 24 "sets" of simulations, and each "set" requires 500-1000 separate (embarrassingly parallel) simulations, each of which runs in 3 GBytes of RAM. Because of the Marvel's outstanding memory bandwidth and MCell's frequent random memory accesses, our simulations run very efficiently even compared to other more recent processors running at higher clock speeds. Perhaps even more important, we have never had any problems related to compilers or operating system issues. This is especially impressive given that each "project" generates up to 48 million output files that would consume up to 2.4 TBytes of disk space, except that we post-process the results on-the-fly, obtaining a reduction of ~1000-fold before transfer to mass storage. Without a stable system combining large memory, outstanding memory bandwidth, fast I/O, and reliable transfer to mass storage, our projects probably could not have been done. Publications: Pattillo, JM, Meriney, SD, and Stiles, JR., 2004, in press, Spatially realistic Monte Carlo simulations predict calcium dynamics underlying transmitter release at a neuromuscular active zone. Soc. Neurosci. Abst. Footnotes: 1. The calcium source is generally more than one channel, each of which is at a different distance from the vesicle that happens to fuse. The calcium sensing (binding) sites are arrayed around the base of each vesicle. The calcium gradient is very steep and different (in space and time) from each channel to each sensor. Thus it is very different from a situation in which multiple binding sites are each responding to the same calcium signal. The apparent cooperativity also depends on how we define the fusion model, e.g., the results are different depending on whether or not we require ~6 sites to be bound simultaneously or just to have been bound at some point in time.